Machine Learning-Driven Structural Analysis of Lifting Self-Forming GFRP Elastic Gridshells

Abstract

The assessment of structural performance is critical for ensuring the safety and longevity of structures built. This study investigates a crucial element of structural damage assessment; the prediction of structural performance for glass fiber reinforced polymer (GFRP) elastic gridshell structures. Utilizing machine learning (ML) methodologies, various algorithms such as linear regression (LR), ridge regression (RR), support vector regression (SVR), K-nearest neighbors (KNN), decision tree (DT), random forest (RF), adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost), category boosting (CatBoost), and light gradient boosting machine (LightGBM) are deployed to forecast the maximum stress. Moreover, first-order and second-order sensitivity analyses are performed to demonstrate the importance of input variables on the output. Through a comparative analysis, CatBoost emerges as the most accurate model, showcasing its potential for enhancing predictive accuracy in structural performance evaluation.